Combining Extreme Learning Machine and Linear Discriminant Analysis for Optimized Apple Leaf Disease Classification

Apple trees are an agricultural commodity with high economic value that often face serious challenges due to various leaf diseases. Early detection and proper treatment are crucial to reducing economic losses and ensuring healthy plant growth. Manual methods for detecting apple leaf diseases tend to be time-consuming and require specialized expertise, making them difficult to apply on a large scale. This study aims to develop an automatic classification system based on image processing to detect apple leaf diseases by combining Extreme Learning Machine (ELM) and Linear Discriminant Analysis (LDA) algorithms. ELM offers high training speed with randomly assigned weights, while LDA performs dimensionality reduction to retain the most relevant features, thereby improving model performance. The study results indicate that the combination of ELM and LDA produces a model capable of classifying apple leaf diseases with an accuracy of 92.50%, an improvement of 6.25% compared to using ELM alone. This demonstrates that the model is highly effective in classifying apple leaf diseases.